[URN] A URN WG Status report

[Leslie Daigle <[e-mail]>]

Hello all,

Many thanks to the folks who contributed info about the URN work that they
have done.  Attached is a survey, or "status report" on the work of this
group and the systems that have been built around this work.

I've outlined which documents remain to be finished (and have bravely
assigned target dates :->  There are a couple of existing RFCs that should
perhaps be re-examined in terms of their status.

Leslie.


----------------------------------------------------------------------------

  "You can lead people to facts, but              Leslie Daigle
       you cannot make them think."               
                                                  Bunyip Information Systems
                      -- ThinkingCat              [phone]
                                                  [e-mail]
----------------------------------------------------------------------------




============================================================
URN Status Report
============================================================
Leslie L. Daigle
IETF URN WG Co-Chair
May 15, 1998.

----------------------------------------
Working Group Documentation
----------------------------------------

The required documentation is just about finished -- minor changes required
to some documents, and one (small) document left to construct.  There
is some discussion about progressing some of the RFCs that have been out
for a year.  Details follow.

Current Internet-Drafts:
-----------------------

   URN Namespace Definition Mechanisms 
      Work required:
	. Concrete references to appropriate "IANA-like" body
	. Minor textual revisions
      Target date:
	. June 1, 1998
	
   URI Resolution Services Necessary for URN Resolution 
      Work required:
	. minor editorial changes, requested by AD
	. resubmission for (experimental) RFC
      Target date:
	. June 1, 1998.

   A URN Namespace for IETF Documents 
      Work required:
	. minor editorial changes (done)
	. final concordance with Namespace Definition Mechanisms
	  paper (if changes necessary)
      Target date:
	. June 15, 1998.

Documents still Required:
------------------------

   NAPTR Registration Procedures document
      Target date:  June 15, 1998.


Request For Comments (RFCs): 
---------------------------

   URN Syntax 
      Status:  Proposed Standard Protocol
      Proposed actions:  ask to move to Draft Standard

   Resolution of Uniform Resource Identifiers using the Domain Name System 
   (RFC2168)
      Status:  Experimental Protocol
      Proposed actions:  ask to move to standards-track; though it may
	not be the best global NID, the proposed implementation seems 
	stable within itself.
     
   A Trivial Convention for using HTTP in URN Resolution (RFC 2169) 
      Status:  Experiemental Protocol
      Proposed action:  none

   Architectural Principles of Uniform Resource Name Resolution (RFC 2276) 
      Status:  Informational
      Proposed action: none

   Using Existing Bibliographic Identifiers as Uniform Resource Names 
   (RFC 2288)
      Status:  Informational
      Proposed action: none


----------------------------------------
Related Work
----------------------------------------

Apart from implementations developed specifically in conjunction with the
URN WG, related initiatives have developed software and/or naming systems
that are complementary to, and potentially with, URNs.


CNRI's Handles:
--------------

Put forward as one of the original proposals for implementing the URN
concept, the Handles technology is being pursued by CNRI as a 

"[...]distributed computer system which stores names, or handles, of digital 
 items and which can quickly resolve those names into the information 
 necessary to locate and access the items. It was designed by CNRI as a 
 general purpose global system for the reliable management of information on 
 networks such as the Internet over long periods of time[...]" 
 (see http://www.handle.net)

The particular implementation choices make this system more applicable for
some needs of URNs than others (hence the decision to pursue a more 
framework-oriented solution to URNs within the IETF).  One of the applications
for which it is being used is the Digital Object Identifier, described below.

See http://www.handle.net/  for more information about CNRI's Handles.



DOIs ("Digital Object Identifiers"):
-----------------------------------

Now supported through the "International DOI Foundation", this is an 
initiative to address the Publishing Industry's needs for identification
and manipulation of electronic documents.  DOIs are still in evolution;
although the syntax and management issues seem to be pretty much fixed, 
the full machinery for supporting digital commerce for the publishing
industry awaits clearer definition of that community's specific requirements.

Formally introduced at the Frankfurt Bookfair in October, 1997, several
publishers have started testing DOIs, and tens of thousands have been
assigned as part of their projects.

The current DOI system is based on CNRI's Handle technology.  A couple of 
sample DOIs are:

Elsevier:
	10.1016/S1384107697000225  which is resolvable through

	http://hdl.handle.net/10.1016/S1384107697000225

Wiley:
	10.1002/0002-8231(199601)47:1<1:SPOTEO>2.3.TX;2-K  which is resolvable
	through

	http://doi.wileynpt.com/10.1002/0002-8231(199601)47:1<1:SPOTEO>2.3.TX;2-K

As I understand it, the "10." identifies the only top-level being used
by the DOI Foundation currently, and the 4 digits afterwards indicate the 
specific publisher which is responsible for the assignment/management of the 
remainder of the DOI.  

The expectation is that DOIs, will be put forward as a URN namespace to allow 
integration with general URI software.  For example,

	urn:doi:10.1016/S1384107697000225 

See http://www.doi.org/  for more details about DOIs and the IDF.



Persistent Document Identifiers (PDIs)

Developed for US government document series, PDIs are intended primarily as 
permanent identifiers for archival reference to long-lived documents, and
tackle many specific problems related to that activity (including 
intracollection fragment references, etc).  They are expected to be made 
available as a URN namespace, and have already been deployed -- see
 
	http://www.pub.whitehouse.gov

to see the system in action.  For example, 

	pdi://oma.eop.gov.us/1993/1/21/1.header.1

is resolved through the proxy:

	http://www.pub.whitehouse.gov/uri-res/I2R?urn:pdi://oma.eop.gov.us/1993/1/21/1.header.1

This system is documented in:

	http://www.ietf.org/internet-drafts/draft-mallery-urn-pdi-00.txt



W3 Identifier Resolution Extensions (WIRE)

This work, still very experimental in nature, explores the possibility of
incorporating resolution extensions directly into the HTTP protocol, through
the use of new redirect mechanisms.

See:

       http://www.ietf.org/internet-drafts/draft-girod-w3-id-res-ext-00.txt
       http://www.ietf.org/internet-drafts/draft-girod-urn-res-using-wire-00.txt

----------------------------------------
Technology
----------------------------------------

Apart from the extensions to BIND to support the NAPTR DNS record, (available
in distributions of BIND for the past couple of years), most of the technology 
that has been developed for URNs has been proprietary and/or prototype in 
nature.  See below for more information on specific prototype projects.

Prototype source code has been made available (see Ryan Moats' IETF namespace
document) that demonstrates the feasibility of using the URN resolution
methods outlined by this group.  

Beyond that, a user-installable software patch will be available (Michael 
Mealling, Network Solutions Inc) to enable Microsoft's Internet Explorer
to handle URIs of the form "URN:". The expectation is that future versions of 
Netscape (the source for which is now available) will incorporate this 
extension.  There is no indication that Microsoft's Internet Explorer won't 
follow suit...


----------------------------------------
Prototype/Deployed Systems
----------------------------------------

Live systems:
------------

"National Bibliography Numbers (NBNs)", Juha Hakala, National Library of Finland 
In the context of creating a large-scale index of web resources (Nordic
Web Index), this work has focused on using NBNs (identifiers assigned
to items that do not have ISBNs, or ISSNs) within URNs to form a permanent
identifier for web documents.  

Currently, the initiative is integrated with the overall Nordic Web Indexing
Project (national web indexes for Nordic countries).  A web page is available
for obtaining a URN for resources within the scope of Finland and Sweden, 
and users are encouraged to use that identifier within the metadata of
their documents so that search services will pick it up.  "Resolution" is 
currently handled through the NWI project's search pages -- i.e., searching 
for a particular URN will yield the "hits" of documents that contain it/refer 
to it.

The NBN namespace is set up to assign 8 000 identifiers this year; the
space will be extended if need exceeds that bound.

The NBN URN generator service went live on May 5, 1998.  The URN-assignment
page for resources in Finland and Sweden is:

	Finnish:
	   http://linnea.helsinki.fi/cgi-bin/urn.pl 
	English:
	   http://www.lub.lu.se/cgi-bin/nmurn.pl

Explanatory information is available from:

	http://www.lub.lu.se/metadata/URN-help.html



Software components:
-------------------

Michael Mealling, Network Solutions Inc ([e-mail])

This is a user-installable extension that enables Microsoft's Internet
Explorer to handle URIs of the form "URN:".  It supports the I2R, I2L,
I2Ls and I2C services (allowing the user to configure whether a dialog
is popped up to select an L for I2Ls, or one is picked at random), and
the thttp, rwhois, and rcds (when stabilized) protocols.

This code has been tested against the thttp server used by Ryan Moats.


Ryan Moats, AT&T ([e-mail])

Perl scripts for I2C, I2L, I2Ls, I2Ns, I2R, I2Rs resolution of the
proposed "IETF" namespace, using the thttp protocol.  This code is 
available as part of the IETF namespace Internet-Draft.

Ryan also has what he describes as a "proof of concept" URN browser
written in perl.



Experimental prototype systems:
------------------------------

Renato Ianella, DSTC Pty ([e-mail])

The DSTC did some work a couple of years ago on a simple URN 
resolver.  It was done before any of the URN documents were really
finalized, so it never got beyond prototype state, but it was
tested for interoperability with Dirk van Gulik's system (no info
currently available); each system could resolve the other's URNs.  
Approximately 30 URNs were assigned, with URC metadata stored in an X.500 
database.

Some of this work is described in a conference paper:

	http://www.dstc.edu.au/RDU/reports/APweb96/index.html

A key lesson from the work was the critical need for browser support
for URNs in order to facilitate use and promotion of the identifiers.

Dirk Willem van Gulik, JRC ([e-mail])

[No info available]



Ron Daniel Jr., Ed Balas, LANL

This focused on the development of a URN library (Ed Balas), which will be
made publicly available, and has already been used as the inspiration for
some of the software tools mentioned above. The software does a first
check for "known" namespaces, falling back to NAPTR if namespaces are not
recognized locally.  Resolution is done through thttp.  Only a few URNs
were assigned/resolved as part of this experimental project.

Key lesson learned:  "Maintenance, maintenance, maintenance".